LEIOGNATHUS 7 



lated to the lower edge of the very short symplectic and the epihyal to 

 the interopercle, as described in greater detail elsewhere in this paper. 



Leiognathus has three distinct tooth-bearing superior pharyngeals 

 on each side about equal in size. Gerres has two : a very small anterior 

 one, and a very large ovoid one behind it, to which the third and fourth 

 arches are very distinctly attached. There are traces of a suture sepa- 

 rating off a small, lunate, posterior portion of this plate, but if this 

 portion represents a third pharyngeal element it is entirely behind the 

 fourth arch and unattached to it. The lower pharyngeals of both genera 

 meet broadly on the median line, but are not at all coalesced. 



There is no very important difference between the shoulder girdles 

 of these two genera. In Gerres the pelvic girdle is wide as in most of 

 the percoid fishes, but in Leiognathus it is deep and compressed, with a 

 wing of bone on each side extending forward so that a deep channel is 

 left between them. This resembles the pelvic girdle of most of the Chae- 

 tonontoid fishes, and of other deep forms related to the Scombroids. 



As might be expected these genera are too closely related to show 

 important differences in the vertebral column. They both have 10 -f- 14 

 vertebra?. The parapophyses in Leiognathus are larger than in Gerres, 

 stand out more prominently, and are unconnected at the base. The last 

 pair are anchylosed, and spread out in a broad trowel-shaped bone about 

 the end of the abdominal cavity. In Gerres they are as in most of the 

 Percoid fishes the last three or four with a bony ridge connecting them 

 in pairs at the bases. 



Internally, as externally, the long anal of Leiognathus is a notice- 

 able difference. The abdominal cavity is very much shorter, with the 

 first interhaemal and haemal bones vertical, instead of sloping far back- 

 wards as in Gerres. 



In Leiognathus there is a very large baseost at the base of each dor- 

 sal and anal ray firmly joined, suturally, to the interspinous bones. In 

 Gerres baseosts are present in connection with these rays, but they are 

 small, detached nodules of bone. 



On the whole these two forms, aside from the similarity between 

 the mouth parts, have no more in common than most any Percoid fish 

 might have with any Scombroid fish. Hence I propose to recognize the 

 Scombroid family Leiognathidoe and the Percoid family Gerridte. 



